Research map / Federated learning and privacy / Personalized and heterogeneous FL
FedNASP: Federated Vision-Language Navigation with Adaptive Step-wise Personalization
Qingqian Yang, Hao Wang, Sai Qian Zhang, Jian Li, Yang Hua, Miao Pan, Tao Song, Zhengwei Qi, Haibing Guan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add SPI (adapter-based personalized prefix injection) (ablation) | FedNASP · REVERIE | SR: 29.83 -> 43.23 | Improved |
| add SPI (adapter-based personalized prefix injection) (ablation) | FedNASP · CVDN | SR: 21.19 -> 39.02 | Improved |
| add SPM (step-wise modulator) to fixed-strength prefix (ablation) | FedNASP · REVERIE | SR: 39.96 -> 43.23 | Improved |
| add SPM (step-wise modulator) to fixed-strength prefix (ablation) | FedNASP · REVERIE | RGS: 30.91 -> 35.22 | Improved |
| add SPM (step-wise modulator) to fixed-strength prefix (ablation) | FedNASP · CVDN | SR: 32.40 -> 39.02 | Improved |
| add SPM (step-wise modulator) to fixed-strength prefix (ablation) | FedNASP · CVDN | SPL: 16.80 -> 23.14 | Improved |
| keep only the language encoder local (personalize language module) (ablation) | FedAvg + local training · REVERIE module study | SR: 27.83 -> 13.23 | Hurt |
| keep only the visual encoder local (ablation) | FedAvg + local training · REVERIE module study | SR: 27.83 -> 17.48 | Hurt |
| keep only the cross-modal module local (ablation) | FedAvg + local training · REVERIE module study | SR: 27.83 -> 33.25 | Improved |
| keep cross-modal, visual and critic modules local (ablation) | FedAvg + local training · REVERIE module study | SR: 27.83 -> 37.35 | Improved |
| add visual to cross-modal personalization (ablation) | cross-modal local · REVERIE module study | SR: 33.25 -> 36.59 | Improved |
| add language to cross-modal personalization (ablation) | cross-modal local · REVERIE module study | SR: 33.25 -> 33.14 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FedPerfix — FedPerfix: Towards Partial Model Personalization of Vision Transformers in Federated Learning (ICCV 2023) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Harmonizing Generalization and Personalization in Federated Prompt Learning (ICML 2024 · similarity 0.87)
- BLEND: Balancing Personalization vs. Generalization in Federated Vision–Language Models (NeurIPS 2026 · similarity 0.87)
- Federated Disentangled Tuning with Textual Prior Decoupling and Visual Dynamic Adaptation (ICML 2025 · similarity 0.87)
- Global and Local Prompts Cooperation via Optimal Transport for Federated Learning (CVPR 2024 · similarity 0.86)
- F^3OCUS - Federated Finetuning of Vision-Language Foundation Models with Optimal Client Layer Updating Strategy via Multi-objective Meta-Heuristics (CVPR 2025 · similarity 0.86)
- FedPHA: Federated Prompt Learning for Heterogeneous Client Adaptation (ICML 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
